Synthetic Administrative Panel Dataset for Performance-Based Rural Road Maintenance Evaluation

This dataset contains a synthetic, simulation-based administrative panel dataset designed to emulate the operational, contractual, and engineering characteristics of a digital performance-based rural road maintenance system. The structure reflects a realistic policy environment in which road maintenance payments are linked to measurable performance thresholds and monitored through a digital inspection platform.

The dataset models 150 rural road assets observed across multiple monthly inspection cycles, creating a balanced panel format suitable for longitudinal, econometric, and machine learning analysis. Each observation represents a road–month inspection record and includes engineering condition scores, payment eligibility status, contractor identity, geographic coordinates, contract duration, and administrative variables.

A digital intervention variable captures staggered adoption of a monitoring platform (e.g., digital scoring or audit system), enabling quasi-experimental evaluation techniques such as:

Difference-in-Differences (DiD)

Event Study Analysis

Regression Discontinuity Design

Instrumental Variables (IV)

Spatial Dependence Models

Dynamic Panel Models

Double Machine Learning (DML)

Performance scoring is based on a composite indicator representing pavement surface condition, drainage quality, shoulders, vegetation clearance, signage, and safety elements. A rule-based incentive mechanism ties payment eligibility to a threshold score of ≥ 80, mimicking real performance-based maintenance contracting (PBMC) environments.

The dataset includes spatial identifiers to support mapping, spillover analysis, treatment diffusion effects, and network-based road governance evaluations. Stochastic variation was introduced to reflect realistic engineering uncertainty, environmental exposure, and measurement noise.

Because the dataset is synthetic, it does not contain personally identifiable information (PII), confidential records, or sensitive administrative content, making it suitable for open publication, reproducibility, benchmarking, teaching, and methodological development in infrastructure policy, econometrics, simulation, and digital governance research.
Key Features
Attribute Details
Type Synthetic, balanced panel dataset
Observations Road × Month structure
Geographic Structure Includes simulated coordinates and spatial adjacency
Intervention Staggered digital monitoring system adoption
Outcome Variables Performance score, payment amount, payment delay
Use Cases Causal inference, econometrics, simulation models, machine learning